Peeling away financial reporting issues one layer at a time

A Loan Accounting Problem to Make “Amortized Cost” Proponents Squirm

Some folks can see fine art in a soup can label, or poetry in a user's manual. I don't mean to brag when I state that I have a similar talent.  I seem to have a knack for stumbling upon interesting accounting issues.

Most recently, I tripped over a New York Times article entitled, "Investors Put Money on Lawsuits to Get Payouts," which describes a new and rapidly growing financing trend: lending to plaintiffs to fund the costs of their lawsuits. Interest rates are astronomical, which explains why everybody seems to be getting into the act, from boutique venture capital firms to behemoths like Citigroup and Deutsche Bank. I wouldn't be surprised if bankers see this innovation in lending like a new drug for repeating the intoxicating highs they got from subprime mortgages.

The article doesn't mention this, but it's important to my analysis.  I have learned through experience as an expert witness that neither lawyers nor their clients may convey a contingent interest in the outcome of a lawsuit to anyone except another lawyer. Thus, as the article states, "Law firms are generally obligated to repay loans even if they lose. In reality, however, firms that make less than expected often struggle to make the required payments, and a number of firms … have filed for bankruptcy protection."

 The Accounting Issues

I'm not going to comment on the social issues of placing bets on lawsuits. Those are the focus of the article itself, and you can find a whole bunch of eloquent readers' comments on all sides of the issues here.  You can read an interesting article written from a legal perspective here.

For accounting purposes, these arrangements fall squarely under the definition of "loan" to be found in the FASB's Accounting Standards Codification: "A contractual right to receive money on demand or on fixed or determinable dates that is recognized as an asset in the creditor's statement of financial position." The lender accounting issues can be illustrated by a simple example, inspired by one of the cases cited in the NYT article:

  • One year ago, a small law firm recruited four hundred residents of one-factory town in Texas to be the plaintiffs for a series of planned lawsuits against the factory owner. A majority of the plaintiffs are being treated for skin cancer, and others for stomach cancer. Scientific research has established a strong link between a chemical used in the manufacturing process (creosote) and skin cancer; the links to stomach cancer are not as strong. All of the lawsuits allege that the cancers are the result of unsafe working conditions. The law firm's litigation strategy is to bring the cases with the lowest probability of success first.
  • To date, the law firm has drawn $3 million from its line of credit provided by single investor at a rate of 20% per annum. Interest is payable annually, the principal amount is due in ten years and early retirement of principal is permitted at no additional cost.
  • The average amount of the loan outstanding at the end of the first year was $1 million, and the law firm made its first interest payment, $200,000.
  • Court filings and other documents made available to the lender indicate that the prospects for awards in the cases have improved since the inception of the line of credit.

Let's consider three accounting alternatives for the lender: (1) current GAAP or IFRS, both of which entail an 'incurred loss' model for measuring loan impairment; (2) proposed IFRS, which is an 'expected loss' model; and (3) proposed GAAP, which would require measurement of outstanding loans at fair value.

Even before we get into loan impairment issues, the first accounting question is the treatment of the $200,000 'interest payment.' As sure as the sun will come up tomorrow, that small law firm must have round-tripped a portion of its drawdown on their line of credit to pay the 'interest,' for it would have had no other source of funds. Yet, current GAAP/IFRS would blithely recognize $200,000 of interest revenue; and it would make no adjustment to the loan balance, since there are no indicators of an incurred loss. If a lender needs short-term earnings — or a swift kick in the pants to its interest rate spread, litigation lending would be a sure fire way to get it done.  And, as crazy as this accounting is, I wouldn't count on any audit firm or issuer to ask the EITF to fix the problem anytime soon.

Obviously, the 'interest revenue' is pure fiction, but explaining why in logical terms could be a challenge. So, here's my attempt at an explanation that will also prove useful as we march forward through the example. The $200,000 is inappropriately based on the "yield to maturity" of 20%, which is in turn based on the contractual payments. In contrast, the economic return (i.e., internal rate of return) is based on the expected payments. One of the reasons why the NYT article inspired me to write this post is that the capital asset pricing model (CAPM) can be used to estimate the economic rate of return of litigation loans pretty well without any complex calculations: since the future returns from the loan are essentially uncorrelated to returns on the "market portfolio," the "beta" of this investment is zero. Therefore, in order to derive the economic value of the loan at any point in time, the capital asset pricing model would suggest that the expected cash flows on the loan should be discounted at the risk-free rate (say, 4%). (Stated another way, the risks of putting money down on a lawsuit are more like the risks of a casino bet with a positive expected value — yes, these do exist, but that's another story — than a corporate investment, because it is fully diversifiable.)

Proposed IFRS would also recognize $200,000 in interest revenue, but then it would give the loan a haircut for the lender's estimate of "expected losses" – despite the known fact that prospects for loan repayment in full had improved during the year. As I have explained in an earlier post, this approach makes no attempt to portray a sensible carrying amount for the loan itself.  Moreover, its ancestors are the secretive income smoothing techniques devised by German bankers that were financially slobberknockered by hyperinflation post WW I. Ich wünsche Ihnen viel Glück, if you are an investor trying to figure out how much the shares of the lender are actually worth from the IFRS financial statements. Good luck, as well, to the auditor risking money and reputation on a lender's biased and unverifiable estimates of the economic effects of unspecified uncertainties.

Proposed GAAP, which has been emphatically rejected by the EU, and consequently the IASB, would require fair value measurement of the loan. There are obvious accuracy issues to be concerned with, but "it is better to be vaguely right than precisely wrong."

At this point, I hope it is already apparent that the only relevant measure of value and earnings under this utterly simple scenario is fair value. And, we haven't even touched on impairment yet. To do that, let's fast forward one more year:

  • The law firm makes the second interest payment on schedule (and from the same source of funds); but two of the weakest cases (involving stomach cancer) went to trial with no damages being awarded to the plaintiffs.

The accounting under current GAAP/IFRS for the second year could be even more ludicrous. First, there is the highly subjective issue of whether or not to recognize an impairment. If yes, then take the lender's latest estimates of expected cash flows and discount them at the original meaningless "yield to maturity" of 20%. More than precisely wrong, it's egregiously wrong – yet, the American Banker's Association has stated that this is the method of loan accounting they like the best. Proposed IFRS is no better. Trust me.

Defenders of amortized cost approaches to loan accounting will say that these litigation loans are not relevant, because they are risky equity investments masquerading as loans. But, if that's true, then junk bonds are also equity. If you can find it within yourself to agree with anything I have written in this post, I want it to be that it is futile for accounting standards to attempt to distinguish between loans and other financial instruments – except for basic ownership interests like common shares or partnership interests in a law firm. The history of financial reporting appears to have persuaded the FASB, but not the IASB, that rules for creating categories of financial instruments beyond this basic dichotomy are utterly ineffective.  Yet, the political farces (pun intended) behind convergence are pushing the FASB away from their principles.  That's not convergence, it's dysfunction.

It didn't take me long after reading the NYT article to realize that litigation lending could be the best example for exposing the absurdity of any amortized cost approach to loan accounting. And, it's not as extreme a scenario as it might initially appear. Too many greed-is-good-style managers have demonstrated an ability and willingness to test the limits of an accounting rule, and there is no one to stop them from saying, "show me the rule that says I can't do this."  Therein lie the seeds of the next multi-trillion financial 'service.' 

And, finally, we shouldn't fall for the story that IFRS's incurred loss model is an incremental improvement over the current expected loss model of amortized cost accounting. It will do nothing to change loan accounting from rules to principles; and it will give lenders greater discretion to massage their numbers.


  1. Reply Independent Accountant December 4, 2010

    Bravo Tom!

  2. Reply Accountants Contractors December 9, 2010

    I was talking to some one in the United States in the past days and what he used to say about the interest rates:
    “Interest is one of those acts which can destroy any business”
    When we talk about any accountancy issue I can say that most of them are due to interest rates.

Leave a Comment